← back to chapters
07 / 54
\part_02Better Together
chapter 07

Robots don't have skin in the game

Bored fox with chin on paw

If you've never killed creative work that tested well, you've never really been accountable for the outcome.

About four years ago, when at The National Lottery, I was looking at System1 research results for our latest campaign. They suggested the campaign would be effective. Not a homerun, but better than the average. Definitely good enough.

But I killed it.

This was a difficult decision. My colleagues were waiting for this campaign. Sales were down. This was a bet we were counting on. I didn't kill it because it wasn't our normal 4 or 5 star ads. I'm a practical marketer. Not everything is a homerun.

What was niggling was some data points that suggested that a small minority would be offended by the advertising. I'm not precious and have been in the game long enough to not worry about keeping everybody happy.

Looking at the research, the ad was still a decent bet. But Irish people feel a certain sense of ownership over The Irish Lottery. It has political oversight. A small minority could create a lot of noise. So I pulled the campaign.

We like to believe that decision-making is rational. Sure, there are frameworks, first principle thinking, and techniques to help us. But we're kidding ourselves if we think we make decisions like robots. For good or bad, we have skin in the game.

Our experiences influence how we see things. They influence our decisions. Pulling that campaign wasn't fun. I went in front of the company, explained my decision, apologised and articulated how I'd fix this. I had to make a decision between how likely that this will blow up, versus how painful is pulling this going to be for me. And decide what is the best outcome for the business.

I use Claude and ChatGPT every day. I share problems I'm working on. Brand challenges. Positioning thinking. I've trained them to use knowledge I use. Frameworks I like. I even have a doc called "How Dervan thinks about strategy" they use. If anybody wants it, let me know.

Claude and ChatGPT are both really useful in this area. They spot problems. They suggest alternatives. They help when I get stuck. They push me if I ask them to. But I can't delegate the decision to them.

They can't feel the weight of the decision. And you need that to make the decision.

#homechapters.md/get_the_book